Left Main Guideline Review Looms as EACTS Surgeons Urge CABG First

Nearly 2 years after the European Society for Cardio-Thoracic Surgery (EACTS) formally withdrew its support for the 2018 guidelines on myocardial revascularization it had co-developed with the European Society of Cardiology (ESC)—specifically the recommendations for left main disease—there may be a light at the end of the tunnel. The chair of the ESC committee on practice guidelines has said that a review of the left main revascularization guidelines should be completed by spring 2022.

Moreover, new follow-up data from relevant trials being considered in the review is just around the corner, Colin Baigent, BMBCh (University of Oxford, England), the ESC committee chair, told TCTMD last week.

As previously reported by TCTMD, EACTS withdrew its support for the current recommendations and called for an independent statistical review of trials underpinning those guidelines after a presentation at the EACTS 2019 meeting where a high-profile cardiac surgeon and former EXCEL trial investigator accused his co-investigators of changing the study’s primary endpoint and burying evidence of harm.

Both ESC and EACTS agreed that the review will include new data published since the guidelines were issued 3 years ago.

“It is intended that the review pays particular attention to longer-term follow-up of the relevant trials, and to a meta-analysis of individual participant data from those trials,” Baigent said in an interview. That meta-analysis is scheduled for presentation November 15, 2021, at the virtual American Heart Association Scientific Sessions. “Hence, ESC and EACTS are now working together to agree on the composition, terms of reference, and a time line for a panel to interpret the results of this meta-analysis and potential implications regarding the recommendations on left main revascularization,” said Baigent.

In addition to the meta-analysis, Baigent said the review will include the 5-year outcomes from EXCEL, as well as the published data on periprocedural MI measured using the Third Universal Definition.

“It should be noted, however, that the panel will be asked to consider the totality of evidence and not just the findings of EXCEL, as it is important that the findings of EXCEL are weighed together with the evidence from other trials,” said Baigent.

The review includes the appointment of a panel of experts to consider all the available evidence in “good faith” and will also include full disclosure of any of the panelists’ conflicts of interest, in addition to the minutes of the panel meetings and a list of all publications that were considered as part of the evidence, said Baigent.   

In the 2018 ESC/EACTS guidelines on myocardial revascularization, both PCI and CABG are class IA recommendations for patients with left main CAD and a low SYNTAX score (0 to 22). CABG surgery is a class IA recommendation for all patients regardless of anatomical complexity, while PCI is a class IIa recommendation in left main CAD patients with an intermediate SYNTAX score (23 to 32), meaning that the “weight of evidence/opinion is in favor of usefulness/efficacy.”

In patients with highly complex anatomy (SYNTAX score ≥ 33), PCI is not recommended.

Debate is Unremitting

Two years after the EXCEL controversy first erupted over the EXCEL trial, debate persists over the best treatment for left main coronary artery disease.

At this year’s EACTS meeting, which wrapped up this weekend in Barcelona, temperatures were ratcheted down a notch or two from 2019, but the fallout over the controversy continues, with several presentations taking aim at the evidence supporting the selected use of PCI for left main CAD.

Cardiac surgeon Victor Dayan, MD, PhD (Universidad de la Republica, Uruguay), an outspoken critic of EXCEL and its leadership, was the opening speaker for a plenary session dedicated to “Evidence and Trial Updates.”

Dayan ran through a host of arguments against PCI for left main CAD, first stating there is “scant evidence” that revascularization decisions should be based on the SYNTAX score. Analyses from EXCEL and NOBLE both showed that PCI was associated with higher mortality and worse clinical outcomes at 5 years, respectively, and there was no interaction between the SYNTAX score and clinical events, he said. In NOBLE at 5 years, which showed CABG was superior to PCI for the primary composite endpoint, Dayan emphasized that patients with a low SYNTAX score (< 23) had a more than twofold higher risk of major adverse cardiovascular and cerebrovascular events. 

Dayan also took aim at the noninferiority margins used in EXCEL, an argument first put forth by Sanjay Kaul, MD (Cedars-Sinai Medical Center, Los Angeles, CA), in the Canadian Journal of Cardiology.

While the main primary endpoint findings in EXCEL are statistically significant, they are “extremely fragile,” said Dayan.

Finally, he noted that if the Third Universal Definition of MI is used as part of EXCEL’s primary composite endpoint—the EXCEL investigators had selected a modified version of the Society for Cardiovascular Angiography and Interventions (SCAI) definition for periprocedural MI and the Universal Definition for spontaneous events—the trial fails to demonstrate noninferiority, showing instead that PCI patients fared worse.

That latter concern has been raised since the trial was presented and published, prompting the EXCEL investigators to explain that they chose the MI definition because it represents an “extensive amount of myonecrosis after both PCI and CABG”—a stance they’ve detailed elsewhere.

Nick Freemantle, PhD (University College London, England), one of the discussants during the EACTS session, didn’t agree with all of the statistical arguments put forth by Dayan, but he admitted that the fight over noninferiority trials can come down to noninferiority margins selected by researchers.

“I think the fundamental problem is that we have a series of trials where people declare remarkably large noninferiority boundaries,” said Freemantle. For example, in their draft guidance for diabetes drugs, the US Food and Drug Administration states that active-controlled trials may use a noninferiority design but the margin in HbA1c reduction must not exceed 0.3%, he noted. Current noninferiority trials in interventional cardiology use much wider margins and are “not excluding no difference,” said Freemantle.  

“To say that both interventions are comparable is misleading,” Dayan concluded, “especially considering that the odds of death at 5 years are 38% higher with PCI.”

John Mandrola, MD (Baptist Health, Louisville, Kentucky), an electrophysiologist, took the “neutral Martian’s” view of the evidence at the end of the plenary. He, too, zeroed in on the difference in mortality at 5 years between PCI and CABG, adding that the mortality curves also appear to be separating. “You can say this was by chance, you can say this was driven by noncardiovascular death, but it is a signal nonetheless,” he stressed.

What About Patients?

Matthias Siepe, MD (University Heart Center, Freiberg, Germany), a cardiothoracic surgeon and panelist during the EACTS plenary, said that while the statistical analysis is important, patients won’t understand the nuance. Physicians need to break the data down for their patients and the bottom line is that there are two clinical trials—EXCEL and NOBLE—both demonstrating better outcomes for surgery at 5 years.

“Right now, I’m a little puzzled, but it’s quite obvious that we can offer a surgical treatment that has better outcomes in terms of mortality after 5 years,” said Siepe. “We offer it to the patient and regardless of the situation, they choose PCI if we offer it to them. Victor, do you have a clue why they do it?”

Dayan replied that all heart teams work differently, and that surgeons often don’t get the opportunity to speak directly with the patient. “In my personal opinion, when we have a common discussion with the cardiologist and the patient with left main [disease], the patient generally chooses surgery,” he said.

Mandrola, seeking some middle ground in his concluding remarks, observed: “For patients similar in characteristics to those enrolled in these trials, CABG should be the preferred approach. But as a clinician who understands that patients are different, I can see in some patients that PCI may be beneficial.”

Another session panelist, Victoria Delgado, MD, PhD (Leiden University Medical Center, the Netherlands), who specializes in noninvasive imaging, said the decision should be left in the hands of the heart team to make the best choice for revascularization, but noted that referring cardiologists often don’t get included in the heart-team discussion. “You get the patient back with the treatment,” she said. As a result, the revascularization strategy for left main disease has come down to a “fight between [surgeons] and interventionalists.”

Mandrola had one other timely reminder for the assembled physicians to close the session. “Cardiac surgery and cardiology have made stunning advances in the past decades, and I’ve been so blessed to be part of this,” he said. “But our patients do best when we work together, not compete against each other.”

Note: Stone and several co-authors of the EXCEL analysis are faculty or employees of the Cardiovascular Research Foundation, the publisher of TCTMD. An earlier version of this story erroneously paraphrased Matthias Siepe, and has been corrected.